1

A NEW CONVERGENT SERIES SOLUTION OF THE BLOCH

NUCLEAR MAGNETIC RESONANCE FLOW EQUATIONS


K.M Fasasi; A.S. Yahya and S.A. Reju
Abstract
The importance of magnetic resonance in classical diagnoses and
prognoses cannot be over emphasized. Among its wide clinical
applications is that it provides accurate information, which is of a
physiological and morphological nature at molecular level, with respect
to the metabolic processes in the human body. It is also useful for the
quantification of the blood flow pattern in the system of human body.
This paper presents the convergent series solution of the Bloch nuclear
magnetic resonance equations for magnetic resonance imaging with
time varying rF B
1
Field.The solution can be used to extract information
in the velocities of interest from the measured MR phase modulated
signal.


Introduction
Magnetic resonance is a phenomena produced by simultaneously
applying a steady magnetic field and electro magnetic radiation (usually radio
waves) to a sample of atoms and then adjusting the frequency of the radiation and
the strength of the magnetic field to produce absorption of the
radiation(Boer,1996).

Nuclear magnetic resonance simply refers to resonance of protons to
radiation in a magnetic field. Nuclear magnetic resonance is analogous to
Electron Paramagnetic Resonance (EPR), however NMR is produced by the
much smaller magnetism associated with impaired nuclear spins. The NMR
resonant frequency (usually that of protons in complex molecules) is slightly
shifted by interaction with nearby atoms in the sample, thus providing
information about the chemical structure of organic molecules and other
materials. NMR is now extensively employed in medicine, although the use of
the world nuclear is avoided, the preferred name being magnetic resonance
imaging. The technique provides high-quality cross- sectional images of internal
organs and structures. The imaging sequence can be modified to visualize blood
flow to compensate for the blurring effects of cardiac or respirator
motion(Makinde,2006). Magnetic resonance also offers the unique ability to
acquire images in virtually any direction, without repositioning the patient. This
translates into greater convenience for medical staff and minimized patient
discomfort. Furthermore, magnetic resonance provides chemical information not
measurable with conventional radiography or ultrasonography. It is the
combination of its versatility, sensitivity and specificity as a diagnostic modality
that has accelerated the acceptance of magnetic resonance imaging.

Nigerian Journal of Research and Production Volume 17 No 2 November, 2010
2

Though, there have been tremendous theoretical and practical advent
improvements of these MR flow techniques, new results of intensive current
researches in this filed are continuously being presented at scientific meetings.
However, a great deal of further research is needed to exhaust all the quantitative
information for studying hemodynanics and study dynamics by magnetic
resonance. An ideal approach to this further research would be to find
generalized time dependent analytical solution to the Bloch NMR flow equations
(Archie,1991).

The Bloch equations are coupled non linear equation describing the
motion of a macroscopic magnetization m under the influence of applied
magnetic fields. There are no simple closed solutions known for a general rF
excitation. Therefore generalized analytical solution of the Bloch NMR flow
equation is obviously a very difficult task.

Awojoyogbe and Salako(2000) used Picards method to obtain the
solution of Bloch Nuclear Magnetic Resonance flow Equations the results
obtained were accurate and stable, but it is bedeviled by some computational
difficulties which make them impracticable. These difficulties include
(i) cumbersome derivation process
(ii) large amount of functions evaluation
(iii) requirement of large amount of time

The method, which we are considering in this paper, reduces the amount
of difficulties and makes it more general in application than present. The
modified method is new convergent series solution of Bloch Equations for NMR
Flow with spatially varying magnetic field gradient.

In this contribution, a convergent series solution of Bloch equations for
NMR. Fluid flow with spatially varying magnetic field gradient is presented.
For this investigation, we assumed that resonance condition exist at larmor
frequency Awojoyogbe(1999)

0 B f
0
= e = (2.1)
is the gyro magnetic ratio of blood spins:

2
is the rF excitation frequency;

0
f is the off resonance filed in the rotating frame of reference. The x,y,z
components ( in the rotating frame) of magnetization of a fluid bolus is given by
the Bloch equation, which may be written as follows.
*
2
T
M
t
M
M grad V
dt
dMx
X X
X
=
c
c
+ =

K.M Fasasi; A.S. Yahya and S.A. Reju

3

T
M
) x ( B M
t
M
M grad * V
dt
dM
2
y
1 x
y
y
y
=
c
c
+ =

T
M M
) x ( B M
t
M
M grad * V
dt
dM
1
z 0
1 x
z
z
z

=
c
c
+ =

Two reasonable initial boundary conditions, which may conform to the
real-time experimental arrangements, are chosen. These are
1.

Z o
M M =

a situation, which holds good in general and in particular when the rF B1 (x)
field is strong, say of the order of 1.0 G or more so that M
z
of the fluid bolus
changes appreciably from M
o.

2. Before entering the signal detection system, blood bolus has magnetization
M
x
=0, M
y
=0 (Rien,1991)
For steady flow 0
t
M
y
=
c
c

From equations (2.2) and (2.3) we obtained:
| | ( ) (
) (
/ 1 / 1
1
1 2
1 0
2
1
2 2
2 1
2
2
T V
x B M
M x
V
B
dx
dM
T T
V dx
M d
y
y y

= + + +
| | ) ( ) ( ) ( ) ( ) ( ) ( ; / 1 / 1
1
1 1
1
2
0
2
1
1
2
2
2
2 1
x B x B
T V
M
x Q and x PB x B
V
x S T T
V
R Setting

= = = = + =
Thus equation (2.5) becomes:
) ( ) (
1
2
1
2 2
2
x B M x PB
dx
RdM
x d
M d
y
y y
= + +
. terms dependent Velocity
T V
M
;
V
P Where
1 2 2
2
0

=
Assume B
1
(x) is a linear function of the form:
B
1
(x ) =ax +b (2.7)
B
1

2
(x ) =a
2
x
2
+2abx +b
2
(2.8)
Setting A =

a
2
; k =2ab and c =b
2
Thus equation (2.6) becomes
d
2
M
y
+RdM
y
+P(Ax
2
+kx +c)M
y
= (ax+b) (2.9)
d
2
x
2
dx
The point x =x
0
is an ordinary point. Thus we have that:
b ax M ) c kx Ax ( P RM M
y
2 I
y y
+ = + + + +
[
(2.10)
Assuming a solution of the form:
A New Convergent Series Solution Of The Bloch Nuclear Magnetic Resonance Flow
Equations

4

r
0 r
r y
x M

=
o = (Makinde,2000) (2.11a)

=
1
1
1
1
. .
r
r
y
x r M e i (2.11b)
and

=
2
2
) 1 (
r
r
r
ii
y
x r r M (2.11c)
Substituting equations (2.11:a,b, c0 into (2.10) to have:
(2.12)
{ } b ax x c x kx x Ax P x r R x ) 1 r ( r
0 r
r
r
0 r 0 r
r
r
r
r
1 r
2 1 r
r
2 r
2 r
r
+ = o + o + o + o + o


=

(2.13)
( ) { } b ax x c x k x A P x r R x r r
r
r
r
r r
r
r
r
r
r r
r
r
r
r
+ = + + + +


=

=

0 0 0
1 2
2 1
1 2
1
(2.14)
Setting r =r-1, r =r-2, r =r-3 and r =r-4 in the 2
nd
,3
rd
, 4
th
and 5
th
terms
respectively. We then have that:
b ax x Pc x Pk x PA x r R x r r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
+ = + + + +


=

2
2
2
3
2
3
4
2
4
2
2
1
2
2
) 1 ( ) 1 (
2.15
Taking two terms in the 1
st
, 2
nd
, and the last summations respectively: and one
term in the 4
th
summation.
b ax x Pc x Pc x Pk x Pk
x PA x r R x R x r r x
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r


+ = + + + + +
+ + + + + +

4
2
2 1 0
4
2
3 0
4
2
4
4
2
1 2 1
4
2
3 2
) (
) 1 ( ) 1 ( ) 1 ( ) 6 2
(2.16)
Combining the like terms together, we have:
x ) a Pc Pk R 2 6 ( ) b Pc R 2
1 0 2 3 0 2
o + o + o + o + o + o + o
{ } 0 x Pc Pk PA ) 1 r ( R ) 1 r ( r
2 r
2 r
4 r
3 r 4 r 1 r r
= o + o + o + o + o +

=


(2.17)

2.1 Recurrence Formula
With coefficients vanishing identically
i.e. 0 b Pc R 2
0 1 2
= o + o + o
(2.18a)
K.M Fasasi; A.S. Yahya and S.A. Reju

5

0 a Pc Pk R 2 6
1 0 2 3
= o + o + o + o
(2.18b)
0 ) 1 ( ) 1 (
2 3 4 1
= + + + +
r r r r r
Pc Pk PA r R r r
(2.18c)
From equation (2.18a) we have that:
2
) Pc R ( b
0 1 2
o + o = o
(2.19)
Similarly from (2.18b) we have also:
6
) Pc Pk R 2 ( a
1 0 2 3
o + o + o = o

Thus substituting
3 2
int o we have:

6
) Pc R ( ) k Rc ( P ) Rb a (
1
2
0 3
o + o + = o
(2.20)
And also from ( 2.18c) we have that:
{ }
) 1 (
) 1 (
1 2 3 4

+ + + =

r r
r R Pc Pk PA
r r r r r

(2.21)
Equation (2.21) Is the Recurrence Formula of o
R
Bloch function.
Hence the solution of equation ( 2.6) is given as:

=
o =
0 r
r
r y
x M
{ }
) 1 r ( r
x ) 1 r ( R Pc Pk PA
r
1 r 2 r 3 r 4 r
0 r

o + o + o + + o
=

(2.22)
From equation (2.21), r takes value from 4, we have the following expressions:
{ }
12
R 3 Pc Pk PA
3 r 2 1 0 4
o + o + o + o = o
(2.23)
{ }
20
R 4 Pc Pk PA
4 3 2 1 5
o + o + o + o = o
(2.24)
{
30
5
5 4 3 2 6
R Pc Pk PA + + + =
(2.25)
{ }
42
R 6 Pc Pk PA
6 5 4 3 7
o + o + o + o = o
(2.26)
{ }
56
R 7 Pc Pk PA
7 6 5 4 8
o + o + o + o = o
(2.27)
{ }
72
R 8 Pc Pk PA
8 7 6 5 9
o + o + o + o = o
(2.28)
Substituting equation (2.19),(2.20)into(2.23) to have:
A New Convergent Series Solution Of The Bloch Nuclear Magnetic Resonance Flow
Equations

6

} | | | | {
2
) ( ) ( ) ( ) ( 12
1
2
0 1 1 0 4
Pc R k Rc P Rb a R Pc R b Pc Pk PA
o
+ + + + + + =

Combination the like terms to have
| | | | { }
24
} ] PK 2 ) Pc 2 R ( R [ )] Pc A 2 ( P ) k Rc ( PR }] Rb a ( R Pcb
1
2
o
2
4
o + + o + + + = o
(2.29)
Putting Equations (2.19),(2.20)and (2.29)into (2.24)
| | | |
| | | | | | | | }
6
2 ) 2 ( ) 2 ( ) ( ) (
6
) ( ) ( ) ( 6 / ) ( 2 / { 20
1
2
0
2
1
2
0 0 1 1 5


PK Pc R R Pc A P k Rc PR Rb A R Pcb R
Pc R k Rc P Rb a Pc Pc R b Pk PA
+ + + + +
+ + + + + =
=
5

{ ( ) | | | |
| | } 120 / 5 6
2 ) ( ) ( 2 ) ( 2 3
1
2 2 2 4
0
2 2 2

c P PcR R PkR PA
RPA c Rk PR k Rc c P Rb a R Rb a Pc Pkb

+ + + +
(2.30)

Similarly: putting (2.19), (2.20), (2.29) and (2.30) into (2.5)


| | | |
| | | | |
| | | | |
| | }| | ) 2 ) 2 ( ) 2 (
) ( ) ( ( 24 / ) 5 6
6
2 ) ( ) ( 2 ) ( ) 2 ( 3 ( 24 /
6
) ( ) ( ) ( 6 / ) ( 2 / { 30
1
2
0
2
1
2 2 2 4
0
2 2 2
1
1
2
0 0 1 6




k P Pc R R Pc A P
k Rc PR Rb a R Pcb Pc c P PcR R PkR PA
A RP c RK PR k Rc c P Rb a R Rb a Pc Pkb R
Pc R k Rc P Rb a Pk Pc R b PA
+ + +
+ + +
+ + +
+ + + + =
Combining like terms to have
{ | | | | | ) 2 ( 3 ) ( ) ( 4 12 720
6
Rb a Pc kPb R Rb a R Pcb Pc Rb a PK Pab + + + =

|| | | ) Pc A 2 ( P ) k Rc ( PR Pc ) k Rc ( k P 4 Ac P 12 ) Rb a ( R
2 2 2 2
+ + +
| || | ( ) Pc R Pk PAR PAR c Rk PR k Rc c P R + +
2
0
2 2
4 12 1 2 ) ( ) ( 2
| | | || }
1
2 2 2 4 2
5 6 2 ) 2 ( PcR c P R PkR PA R Pk Pc R R Pc +
Hence we have:
K.M Fasasi; A.S. Yahya and S.A. Reju

7

| { | ) ( 3 ) 2 ( ) 7 4 ( 12 : .
3 2
6
Rb a R bPc R Ra Pcb Pc Rb a Pk Pab + + + + =

|
0
3 4 3 3 2 2 2 2 2 2 2
3 4 2 ) 2 ( 7 c PR k PR c P R k P k P a PR kR A c P + + + + + +
| | } 720 / ) R 2 ( c P R kc P 6 PAR 18 PkR 9
1
2 2 5 2 2
o + + + +
Thus equation ( 2.22) becomes:
3
y
2
y
1
y y
M M M M + + =
Where:
|
} +

+
+
+

+ =
5
2 4
3
2
120
) ( ) (
3
24
)) (
( )
6
(
2
x
rb a R Rb a Pc b
kP
x Rb a R Pcb
x
Rb a bx
M
y

) | |

+ +
=
24
) 2 ( ) (
6
(
2
4 2 3 2 2
2
x Pc A P k Rc PR x k Rc P x Pc
M
y

| |
0
5 2 2
120
x RPA 2 ) c Rk ( PR ) k Rc ( c P 2
o
+ +
+
{
}

+
(

+
+ +
=
1
5
2 2 2 4
4
2 3 2 2
3
120
5 6
24
2 ) 2
(
6
) (
2
x
c P PcR R kR
P PA x
Pk Pc R
R
x Pc R Rx
M
y
3 y
M and M
2 y
can be confidently neglected, because they do not contain the
term M
0
where thus.
Also by neglecting all other terms embedded in term . We thus have the below
expression:
{ |
}

+ = =
720
x ) Rb A (
R
x
120
) rB A
( R
24
x ) Rb A ( R
6
x ) Rb a (
2
bx
M M
6 3 5
2
4 3 2
1 y
y
(2.31)
From equation (2.31) above, we have the general formula as
( )

=
+

+
+
=
2
1
2
!
)
)! 1 (
( ) 1 (
n
n n
n n
y
n
bx
n
ax
R M
(2.32)
For b =0 we have that:
)
`

+ +

=
7
4
6
3 5 2
4
3
! 7 ! 6 ! 5 ! 4 ! 3
aX R ax R ax R Rax ax
M
y

=
+

+
=
2
1
) 2 (
)! 1 (
) 1 (
n
n
n n
y
n
X
R a M
(2.33)

A New Convergent Series Solution Of The Bloch Nuclear Magnetic Resonance Flow
Equations

8

Conclusion
A cursory look at the derived solution shows that this method is fast and
more direct than the Picards method already used by previous researchers. The
new convergent series method is particularly attractive because it does not
require large amount of functions evaluation and therefore could be used as
formula in a multipurpose code for analysis of real life problems arising in
medical physics, medicine and other applied sciences.

References
Archie J .P (1991). Haemodynamics analysis of a stenosed carotid bifurcation and
its plaque-mitigating design. Journal of biomechanical engineering 58(2),
113 335.

Awojoyogbe O.B. (1999): Analytical Solution of the Bloch Equation for MRI
Flow Dependent Magnetization and Signal with spatial varying rF B
1
(x)
Magnetic field. Jour. Nig. Ass. Math. Phys. 3, 360-369.

Awojoyogbe, O.B . (2002). A mathematical model of Nuclear Magnetic
Resonance Equations for quantitative analysis of blood flow in blood
vessels with changing cross- section. Physical 303, 163-175

Boer, J .A. (1996). Magnetic resonance imaging: theory and practice. Berlin
Heidelberg: New York.

Makinde, O.D. (1999). Hemodynamics of the retinal aneurysms: A Mathematical
model. Journal Fizik malaysia 20(1), 43-58

Makinde, O.D. (2000a) . Numerical investigation of bifurcation in Visco-Elastic
flow problem. Zimbabwe Journal of Science and technology,1(1), 10-15.

.Makinde, O.D. (2000b). Effect of variable viscosity on arterial blood flow.
Journal Appl. Maths 4(1), 43-58.

Rieu, R. (1991). In Vitro Study of physiological type flow in a bifurcation
vascular prosthesis. Journal of Bio mechanics 24 ( 1), 923-933.

Salako,K.A.(2000). Studies on polynomial solution of Bloch NMR flow equation
for the estimation of blood flow rates. Physical 202, 141-148.

K.M Fasasi; A.S. Yahya and S.A. Reju